search

chill117 / lnurl-node

Flexible lnurl server implementation with API and CLI implemented in nodejs.
MIT License
97 stars 26 forks source link

readme

lnurl-node

Build Status

Node.js implementation of lnurl. The purpose of this project is to provide an easy and flexible lnurl server that you can run as a stand-alone process (via CLI) or integrated with your own custom node.js application (via API).

Optionally, your lnurl server can authorize other applications (offline or otherwise). Possible use-cases include offline Lightning Network ATMs (e.g. Bleskomat), static QR codes for receiving donations, authentication mechanism for web sites or web services (login / registration / 2FA).

This project attempts to maintain backwards compatibility for any features, methods, options, hooks, and events which are documented here.

Specification Support

The LNURL specification is divided into separate documents called "LUDs". These documents can be found in the lnurl-rfc repository.

The following is a list of LUDs which this module already (or plans to) support:

Installation

If you wish to use this module as a CLI tool, install it globally via npm:

npm install -g lnurl

Add to your application via npm:

npm install lnurl --save

This will install lnurl and add it to your application's package.json file.

Command-line interface

This section assumes that you have lnurl installed globally and that it is available on your current user's PATH.

CLI: help

To view the help menu:

lnurl --help

CLI: encode

Encode a URL:

lnurl encode "https://service.com/api?q=3fc3645b439ce8e7f2553a69e5267081d96dcd340693afabe04be7b0ccd178df"

Expected output:

lnurl1dp68gurn8ghj7um9wfmxjcm99e3k7mf0v9cxj0m385ekvcenxc6r2c35xvukxefcv5mkvv34x5ekzd3ev56nyd3hxqurzepexejxxepnxscrvwfnv9nxzcn9xq6xyefhvgcxxcmyxymnserxfq5fns

This command also accepts piped input. For example:

echo -n "https://service.com/api?q=3fc3645b439ce8e7f2553a69e5267081d96dcd340693afabe04be7b0ccd178df" \
    | lnurl encode

CLI: decode

Decode an lnurl:

lnurl decode "lnurl1dp68gurn8ghj7um9wfmxjcm99e3k7mf0v9cxj0m385ekvcenxc6r2c35xvukxefcv5mkvv34x5ekzd3ev56nyd3hxqurzepexejxxepnxscrvwfnv9nxzcn9xq6xyefhvgcxxcmyxymnserxfq5fns"

Expected output:

https://service.com/api?q=3fc3645b439ce8e7f2553a69e5267081d96dcd340693afabe04be7b0ccd178df

This command also accepts piped input. For example:

echo -n "lnurl1dp68gurn8ghj7um9wfmxjcm99e3k7mf0v9cxj0m385ekvcenxc6r2c35xvukxefcv5mkvv34x5ekzd3ev56nyd3hxqurzepexejxxepnxscrvwfnv9nxzcn9xq6xyefhvgcxxcmyxymnserxfq5fns" \
    | lnurl decode

CLI: generateNewUrl

To generate a new lnurl that a client application can then use:

lnurl generateNewUrl \
    --host "localhost" \
    --port "3000" \
    --endpoint "/lnurl" \
    --store.backend "knex" \
    --store.config '{"client":"postgres","connection":{"host":"127.0.0.1","user":"postgres","password":"example","database":"lnurl_example"}}' \
    --tag "withdrawRequest" \
    --params '{"minWithdrawable":10000,"maxWithdrawable":10000,"defaultDescription":""}'

See Tags and Parameters for a full list of possible tags and params.

Alternatively, a configuration file can be used:

lnurl generateNewUrl \
    --configFile ./config.json \
    --tag "withdrawRequest" \
    --params '{"minWithdrawable":10000,"maxWithdrawable":10000,"defaultDescription":""}'

Example output:

{
    "encoded": "lnurl1dp68gup69uhkcmmrv9kxsmmnwsarxvpsxqhkcmn4wfkr7ufavvexxvpk893rswpjxcmnvctyvgexzen9xvmkycnxv33rvdtrvy6xzv3ex43xzve5vvexgwfj8yenxvm9xaskzdmpxuexywt9893nqvcly0lgs",
    "secret": "c2c069b882676adb2afe37bbfdb65ca4a295ba34c2d929333e7aa7a72b9e9c03",
    "url": "http://localhost:3000/lnurl?q=c2c069b882676adb2afe37bbfdb65ca4a295ba34c2d929333e7aa7a72b9e9c03"
}

It is possible to set the number of uses allowed for the new URL:

lnurl generateNewUrl \
    --configFile ./config.json \
    --tag "withdrawRequest" \
    --params '{"minWithdrawable":10000,"maxWithdrawable":10000,"defaultDescription":""}' \
    --uses 3

Set --uses equal to 0 to allow the URL to be used an unlimited number of times.

For a list of available options:

lnurl generateNewUrl --help

It is also possible to generate lnurls in other ways:

CLI: server

Start an lnurl application server with the following command:

lnurl server \
    --host "localhost" \
    --port "3000" \
    --lightning.backend "dummy" \
    --lightning.config '{}'

Alternatively, a configuration file can be used:

lnurl server --configFile ./config.json

To print all available options for the server command:

lnurl server --help

API

encode

encode(url)

Encode a url as a bech32-encoded string.

Usage:

const lnurl = require('lnurl');
const encoded = lnurl.encode('https://service.com/api?q=3fc3645b439ce8e7f2553a69e5267081d96dcd340693afabe04be7b0ccd178df');
console.log(encoded);

Expected output:

"lnurl1dp68gurn8ghj7um9wfmxjcm99e3k7mf0v9cxj0m385ekvcenxc6r2c35xvukxefcv5mkvv34x5ekzd3ev56nyd3hxqurzepexejxxepnxscrvwfnv9nxzcn9xq6xyefhvgcxxcmyxymnserxfq5fns"

decode

decode(url)

Decode a bech32-encoded lnurl.

Usage:

const lnurl = require('lnurl');
const decoded = lnurl.decode('lnurl1dp68gurn8ghj7um9wfmxjcm99e3k7mf0v9cxj0m385ekvcenxc6r2c35xvukxefcv5mkvv34x5ekzd3ev56nyd3hxqurzepexejxxepnxscrvwfnv9nxzcn9xq6xyefhvgcxxcmyxymnserxfq5fns');
console.log(decoded);

Expected output:

"https://service.com/api?q=3fc3645b439ce8e7f2553a69e5267081d96dcd340693afabe04be7b0ccd178df"

createServer

createServer([options])

Create and initialize an instance of the lnurl server.

Usage:

const lnurl = require('lnurl');
const server = lnurl.createServer({
    host: 'localhost',
    port: 3000,
    auth: {
        apiKeys: [
            {
                id: '46f8cab814de07a8a65f',
                key: 'ee7678f6fa5ab9cf3aa23148ef06553edd858a09639b3687113a5d5cdb5a2a67',
                encoding: 'hex',
            },
        ],
    },
    lightning: {
        backend: 'dummy',
        config: {},
    },
});

createServer: options

Below is the full list of options that can be passed to the createServer method.

{
    // The host for the web server:
    host: 'localhost',
    // The port for the web server:
    port: 3000,
    // Whether or not to start listening when the server is created:
    listen: true,
    // Whether or not to allow CORS requests:
    cors: false,
    // The URL where the server is externally reachable (e.g "https://your-lnurl-server.com"):
    url: null,
    // The URI path of the web API end-point:
    endpoint: '/lnurl',
    // See list of possible LN backends here:
    // https://github.com/chill117/lnurl-node#supported-lightning-network-backends
    lightning: {
        // The name of the LN backend to use:
        backend: 'dummy',
        // Configuration options to pass to LN backend:
        config: {},
    },
    store: {
        // Name of store backend ('knex', 'memory'):
        backend: 'memory',
        // Configuration options to pass to store:
        config: {},
    },
    payRequest: {
        // A number greater than 0 indicates the maximum length of comments.
        // Setting this to 0 ignores comments.
        //
        // Note that there is a generally accepted limit (2000 characters)
        // to the length of URLs; see:
        // https://stackoverflow.com/questions/417142/what-is-the-maximum-length-of-a-url-in-different-browsers/417184#417184
        //
        // Since comments are sent as a query parameter to the callback URL,
        // this limit should be set to a maximum of 1000 to be safe.
        commentAllowed: 500,
        // Default metadata to be sent in response object:
        metadata: '[["text/plain", "lnurl-node"]]',
    },
}

generateNewUrl

generateNewUrl(tag, params)

To generate a new lnurl that a client application can then use:

const tag = 'payRequest';
const params = {
    minSendable: 10000,
    maxSendable: 200000,
    metadata: '[["text/plain", "lnurl-node"]]',
    commentAllowed: 500,
};
server.generateNewUrl(tag, params).then(result => {
    const { encoded, secret, url } = result;
    console.log({ encoded, secret, url });
}).catch(error => {
    console.error(error);
});

Expected output:

{
    "encoded": "lnurl1dp68gup69uhkcmmrv9kxsmmnwsarxvpsxqhkcmn4wfkr7ufavvexxvpk893rswpjxcmnvctyvgexzen9xvmkycnxv33rvdtrvy6xzv3ex43xzve5vvexgwfj8yenxvm9xaskzdmpxuexywt9893nqvcly0lgs",
    "secret": "c2c069b882676adb2afe37bbfdb65ca4a295ba34c2d929333e7aa7a72b9e9c03",
    "url": "http://localhost:3000/lnurl?q=c2c069b882676adb2afe37bbfdb65ca4a295ba34c2d929333e7aa7a72b9e9c03"
}

See Tags and Parameters for a full list of possible tags and params.

It is possible to set the number of uses allowed for the new URL:

const tag = 'payRequest';
const params = {
    minSendable: 10000,
    maxSendable: 200000,
    metadata: '[["text/plain", "lnurl-node"]]',
    commentAllowed: 500,
};
const options = {
    uses: 3,
};
server.generateNewUrl(tag, params, options).then(result => {
    const { encoded, secret, url } = result;
    console.log({ encoded, secret, url });
}).catch(error => {
    console.error(error);
});

Set uses equal to 0 to allow the URL to be used an unlimited number of times.

It is also possible to generate lnurls in other ways:

Tags and Parameters

Below you will find all tags and their associated params.

channelRequest:

name type notes
localAmt integer (sats) > 0
pushAmt integer (sats) <= localAmt

login:

none

payRequest:

name type notes
minSendable integer (msats) > 0
maxSendable integer (msats) >= minSendable
metadata string stringified JSON
commentAllowed integer character limit for comments (max. 1000), set to 0 to disallow comments

withdrawRequest:

name type notes
minWithdrawable integer (msats) > 0
maxWithdrawable integer (msats) >= minWithdrawable
defaultDescription string

Hooks

It is possible to further customize your lnurl server by using hooks to run custom application code at key points in the server application flow.

How to use a hook:

const lnurl = require('lnurl');
const server = lnurl.createServer();
const { HttpError } = require('lnurl/lib');

// The callback signature can vary depending upon the hook used:
server.bindToHook('HOOK', function(arg1, arg2, arg3, next) {
    // Fail the request by calling next with an error:
    next(new Error('Your custom error message'));
    // Use the HttpError constructor to pass the error to the response object:
    next(new HttpError('Custom error sent in the response object', 400/* status code */));
    // Or call next without any arguments to continue with the request:
    next();
});

Hook: error

This hook is called when any request fails with an error.

Override the existing error with a new error:

server.bindToHook('error', function(error, req, res, next) {
    // Call next with a new error which will be passed to the default error handler:
    next(new HttpError('A new error message', 400));
});

Continue with the existing error:

server.bindToHook('error', function(error, req, res, next) {
    // Call next with no arguments to let the error pass-thru to the default error handler:
    next();
});

Make a custom response:

server.bindToHook('error', function(error, req, res, next) {
    // Use the response object to make a custom response:
    res.status(200).json({ message: 'custom error handling' });
    // Do not call next in this case.
});

Hook: status

This hook is called when a request is received at the web server's /status end-point.

Example OK status response:

server.bindToHook('status', function(req, res, next) {
    // Call next() with no arguments to continue with the normal status flow.
    next();
});

Example failed status response:

server.bindToHook('status', function(req, res, next) {
    // Or, call next with an error to fail the request:
    next(new HttpError('Service temporarily unavailable', 503));
});

Hook: url:process

This hook is called when a request is received at the web server's LNURL end-point.

Example modifying the request object:

server.bindToHook('url:process', function(req, res, next) {
    req.query.defaultDescription = 'custom default description';
    next();// Call next() with no arguments to continue the request.
});

Example rejecting the request:

server.bindToHook('url:process', function(req, res, next) {
    // Call next() with an error to fail the request:
    next(new HttpError('Failed check', 400));
});

Hook: login

login

The lnurl-auth subprotocol allows users to login/authenticate with your service. You can use the login hook as shown here to execute your own custom code whenever there is a successful login/authentication attempt for your server.

server.bindToHook('login', function(key, next) {
    // This code is executed when the lnurl-auth checks have passed (e.g valid signature provided).
    // `key` is the public linking key which has just authenticated.
    // Perform asynchronous code such as database calls here.
    // Call next() without any arguments to continue with the request:
    next();
});

Hook: channelRequest:validate

Hook: payRequest:validate

Hook: withdrawRequest:validate

channelRequest:validate payRequest:validate withdrawRequest:validate

These hooks are called when validating the parameters provided when creating a new URL. For example, when calling server.generateNewUrl(tag, params).

server.bindToHook('channelRequest:validate', function(params, next) {
    // Throw an error to prevent the creation of the new URL:
    next(new Error('Invalid params!'));
    // Call next() without any arguments to continue with the creation of the new URL:
    next();
});

Hook: channelRequest:info

Hook: payRequest:info

Hook: withdrawRequest:info

channelRequest:info payRequest:info withdrawRequest:info

These hooks are called when the initial request is made to the LNURL end-point. The initial request occurs when a wallet app first scans a QR code containing an LNURL. The wallet app makes the initial request for more information about the tag and other parameters associated with the LNURL it just scanned.

server.bindToHook('channelRequest:info', function(secret, params, next) {
    // `secret` is the k1 value that when hashed gives the unique `hash`
    //  associated with an LNURL in the data store.
    // `params` are the parameters provided when the URL was created.
    // Throw an error to fail the request:
    next(new HttpError('Custom error sent in the response object', 400/* status code */));
    // Call next() without any arguments to continue with the request:
    next();
});

Hook: channelRequest:action

Hook: payRequest:action

Hook: withdrawRequest:action

channelRequest:action payRequest:action withdrawRequest:action

These hooks are called when the second request is made to the LNURL end-point. This request occurs when the wallet app wants to complete the action associated with the LNURL it scanned and made an initial request for previously.

server.bindToHook('channelRequest:action', function(secret, params, next) {
    // `secret` is the k1 value that when hashed gives the unique `hash`
    //  associated with an LNURL in the data store.
    // `params` are the parameters provided when the URL was created plus
    // the parameters provided in the request to the server.
    // Throw an error to fail the request:
    next(new HttpError('Custom error sent in the response object', 400/* status code */));
    // Call next() without any arguments to continue with the request:
    next();
});

Note that these hooks are executed before the server calls the LN backend method. So if an error is thrown here, a channel will not be opened; a new invoice will not be generated; the provided invoice will not be paid.

Events

The server object extends from the event emitter class. It is possible to listen for events as follows:

const lnurl = require('lnurl');
const server = lnurl.createServer();

server.on('EVENT', function(event) {
    // The event object varies depending upon the event type.
});

Event: login

This event is emitted after a successful login attempt.

server.on('login', function(event) {
    const { key, hash } = event;
    // `key` - the public key as provided by the LNURL wallet app
    // `hash` - the hash of the secret for the LNURL used to login
});

Event: channelRequest:action:processed

This event is emitted after a successful call to the LN backend's openChannel method.

server.on('channelRequest:action:processed', function(event) {
    const { secret, params, result } = event;
    // `result` is the non-normalized response object from the LN backend
    // So this will vary depending upon the backend used.
});

Event: payRequest:action:processed

This event is emitted after a successful call to the LN backend's addInvoice method.

server.on('payRequest:action:processed', function(event) {
    const { secret, params, result } = event;
    const { id, invoice } = result;
    // `id` - non-standard reference ID for the new invoice, can be NULL if none provided
    // `invoice` - bolt11 invoice
});

Event: withdrawRequest:action:processed

This event is emitted after a successful call to the LN backend's payInvoice method.

server.on('withdrawRequest:action:processed', function(event) {
    const { secret, params, result } = event;
    const { id } = result;
    // `id` - non-standard reference ID for the payment, can be NULL if none provided
});

Event: channelRequest:action:failed

This event is emitted after a failed call to the LN backend's openChannel method.

server.on('channelRequest:action:failed', function(event) {
    const { secret, params, error } = event;
    // `error` - error from the LN backend
});

Event: payRequest:action:failed

This event is emitted after a failed call to the LN backend's addInvoice method.

server.on('payRequest:action:failed', function(event) {
    const { secret, params, error } = event;
    // `error` - error from the LN backend
});

Event: withdrawRequest:action:failed

This event is emitted after a failed call to the LN backend's payInvoice method.

server.on('withdrawRequest:action:failed', function(event) {
    const { secret, params, error } = event;
    // `error` - error from the LN backend
});

Supported Lightning Network Backends

See lightning-backends for a list of supported Lightning Network backends and their corresponding configuration options.

Configuring Data Store

By default the lnurl server will store data in memory - which is not ideal for several reasons. It is strongly recommended that you configure a proper data store for your server. This module supports PostgreSQL.

PostgreSQL

To use PostgreSQL as your data store you will need to install the postgres module and knex wherever you are running your lnurl server:

npm install knex pg

Then you can run your server via the API as follows:

const lnurl = require('lnurl');
const server = lnurl.createServer({
    // ...
    store: {
        backend: 'knex',
        config: {
            client: 'postgres',
            connection: {
                host: '127.0.0.1',
                user: 'lnurl_server',
                password: '',
                database: 'lnurl_server',
            },
        },
    },
    // ...
});

Or via the CLI:

lnurl server \
    --store.backend="knex" \
    --store.config='{"client":"postgres","connection":{"host":"127.0.0.1","user":"lnurl_server","password":"","database":"lnurl_server"}}'

Debugging

This module uses debug to output debug messages to the console. To output all debug messages, run your node app with the DEBUG environment variable:

DEBUG=lnurl* node your-app.js

Or if using the CLI interface:

DEBUG=lnurl* lnurl server

Tests

To run all tests:

npm test

Changelog

See CHANGELOG.md

License

This software is MIT licensed:

A short, permissive software license. Basically, you can do whatever you want as long as you include the original copyright and license notice in any copy of the software/source. There are many variations of this license in use.